Hi, I am a computer nerd. I also took a computer programming class and got the highest score in the class, but I never followed up with advanced classes. Recently, I’ve thought of different ideas for software I’d like to try to create. I’ve heard about vibe coding. I know real programmers make fun of it, but I also have heard so much about it and people using it and paying for it that I have a hard time believing it writes garbage code all the time.
However, whenever I am trying to do things in linux and don’t know how and ask an LLM, it gets it wrong like 85% of the time. Sometimes it helps, but a lot of times it’s fucking stupid and just leads me down a rabbit hole of shit that won’t work. Is all vibe coding actually like that too or does some of it actually work?
For example, I know how to set up a server, ssh in, and get some stuff running. I have an idea for an App and since everyone uses smart phones (unfortunately), I’d probably try to code something for a smart phone. But would it be next to impossible for someone like me to learn? I like nerdy stuff, but I am not experienced at all in coding.
I also am not sure I have the dedication to do hours and hours of code, despite possible autism, unless I were highly fucked up, possibly on huge amounts of caffeine or microdosing something. But like, it doesn’t seem impossible.
Is this a rabbit hole worth falling into? Do most Apps just fail all the time? Is making an App nowadays like trying to win a lotto?
It would be cool to hear from real App developers. I am getting laid off, my expenses are low because I barely made anything at my job, I’ll be getting unemployment, and I am hoping I can get a job working 20-30 hours a week and pay for my living expenses, which are pretty low.
Is this a stupid idea? I did well in school, but I’m not sure that means anything. Also, when I was in the programming class, the TA seemed much, much smarter at programming and could intuitively solve coding problems much faster due to likely a higher IQ. I’m honestly not sure my IQ is high enough to code. My IQ is probably around 112, but I also sometimes did better than everyone on tests for some reason, maybe because I’m a nerd. I’m not sure I will have the insight to tackle hard coding problems, but I’m not sure if those actually occur in real coding.
No, making an app is not just something you can decide you want to do and do it without learning to code.
If you “vibe code” your way through trial and error to an app, it may work.
But if you don’t understand what it’s doing, why it’s doing it and how it’s doing it?
Then you can’t (easily) maintain it.
If you can’t fix bugs or add features, you don’t have a saleable product - you have a proof of concept.AI tools are useful, but letting the tool do all the driving is asking for the metaphorical car to crash.
probably useful in getting your resume to be noticed?
I use it as a friendlier version of stackoverflow. I think you should generally know / understand what you are doing because you have to take everything it says with a grain of salt. It’s important to understand that these assistants can’t admit that they don’t know something and come up with random generated bullshit instead so you can’t fully trust their answers.
So you still need to understand the basics of software development and potential issues otherwise it’s just negligence.
On a general note: IQ means nothing. I man a lot of IQ tests use pattern recognition tasks that can be helpful but still, having a high IQ says nothing about you ability as developer
On a general note: IQ means nothing. I mean a lot of IQ tests use pattern recognition tasks that can be helpful but still, having a high IQ says nothing about you ability as developer
to put this another way… expertise is superior to intelligence. Unfortunately we have this habit of conflating the two. intelligent people some times do some incredibly stupid things because they lack the experience to understand why something is stupid.
Being a skilled doctor or surgeon doesn’t make you skilled at governance. two different skillsets.
It’s cool for little things when working on an unfamiliar project or learning something new.
But don’t trust one example and read about the features you’re using.
Think of LLMs as the person who gets good marks in exams because they memorized the entire textbook.
For small, quick problems you can rely on them (“Hey, what’s the syntax for using rsync between two remote servers?”) but the moment the problem is slightly complicated, they will fail because they don’t actually understand what they have learnt. If the answer is not present in the original textbook, they fail.
Now, if you are aware of the source material or if you are decently proficient in coding, you can check their incorrect response, correct it, and make it your own. Instead of creating the solution from scratch, LLMs can give you a push in the right direction. However, DON’T consider their output as the gospel truth. LLMs can augment good coders, but it can lead poor coders astray.
This is not something specific to LLMs; if you don’t know how to use Stackoverflow, you can use the wrong solution from the list of given solutions. You need to be technically proficient to even understand which one of the solutions is correct for your usecase. Having a strong base will help you in the long run.
The main problem with LLMs is that they’re the person who memorised the textbook AND never admit they don’t know something.
No matter what you ask, an LLM will give you an answer. They will never say “I don’t know”, but will rather spout 100% confident bullshit.
The “thinking” models are a bit better, but still have the same issue.
No matter what you ask, an LLM will give you an answer. They will never say “I don’t know”
There is a reason for this. LLMs are “rewarded” (just an internal scoring mechanism) for generating an answer. No matter what you say, it will try to maximize the reward value by generating an answer with high hallucination. There is no reward mechanism for saying “I don’t know” to a difficult question.
I am not into research on LLMs, but i think this is being worked upon.
Something very similar is also true with humans. People just love to have answers even if they aren’t entirely reliable or even true. Having just some answer seems to be more appealing than not having any answers at all. Why do you think people had weird beliefs about stars, rainbows, thunder etc.
The way LLMs hallucinate is also a little weird. If you ask about quantum physics things, they actually can tell you that modern science doesn’t have a conclusive answer to your question. I guess that’s because other people have written articles about the very same question, and have pointed out that it’s still a topic of ongoing debate.
If you ask about robot waitresses used in a particular restaurant, it will happily give you the wrong answer. Obviously, there’s not much data about that restaurant, let alone any academic debate, so I guess that’s also reflected in the answer.
It already kinda works 95% of the way. But more often than not the last 5% still requires you to understand everything the AI did which can be hard. If it was you implementing everything, you’d already have the whole context in working memory. I’ve been learning better prompting and getting better at it. I think it thrives in typed languages and where the code base has clear design patterns it can follow.
Only for really basic things. I’m trying to use it to build tools in the background while I do real work, but it quickly falls into a pattern of presenting a working product that actually doesn’t work at all.
If you care about it at all, don’t vibe it otherwise go hog wild
if you know how to code, you can vibe code because you can immediately see and be confident enough to identify and not use obvious mistakes, oversights, lack of security, and missed edge cases the LLM generated.
if you don’t know how to code, you can’t vibe code, because you think the LLM is smarter than you and you trust it.
Imagine saying “I’m a mathematician” because you have a scientific calculator. If you don’t know the difference between RAD and DEG and you just start doing calculations without understanding the unit circle, then building a bridge based on your math, you’re gonna have a bad time.
Nah, building will be fun. Bad time will be for those who will use that bridge:)
I have successfully written and deployed a number of large complex applications with 100% AI written code, but I micromanage it. I’ve been developing software for 30 years and use AI as a sort of code paintbrush. The trick is managing the AI context window to keep it big enough to understand its task but small enough to not confuse it.
Out of curiosity, what do they do and did you use agentic approach or prompt and then copy?
One library I made is https://github.com/pay-theory/dynamorm This is a library to make DynamoDB easier to use in Go, it is used alongside https://github.com/pay-theory/lift which provides similar utility for lambdas
Both of these are in active use by our team and were created using Claude code and cursor.
Please don’t upvote this person, I think they’re a bot. The libs use AWS SDK internally and claim 90% performance boost over AWS SDK, and the person explains it as “you can write less verbose code so development time is shorter”
Thank you.
How did you check the performance though for the ORM? You claim it’s faster that AWS SDK, which literally impossible, as you are using AWS SDK to power it.
The code loads faster when using the pen because the code footprint is smaller, DynamoDB in AWS sdk is very verbose, using the library means that verbosity is reduced significantly as you incorporate more tables and indexes.
It’s been load tested against code using DynamORM and not using it.
The point is not a few less milliseconds, it’s many hours of reduced development for people implementing DynamoDB
The point is not a few less milliseconds, it’s many hours of reduced development for people implementing DynamoDB
So you’re comparing claimed performance (execution) gains to development time? Yeah, that makes sense.
I think you’re a bot.
I pretty much only use it to generate boilerplate. I’ve tried using it to learn the syntax of new languages and it kind of works, but in my experience just reading the docs is better even if it seems like a lot of text. Also your IQ really does not matter. You can learn anything as long as you’re willing to put in the time and effort; don’t compare yourself to others it’s fine to go at your own pace. (I’m Autistic also btw)
Concerning the IQ: App development and regular programming aren’t that hard. It needs some time and dedication, and willingness to learn how all these things work and tie together, but I think everyone with an average IQ could do it. It’s specific domains where you need a high IQ, like writing advanced signal processing algorithms. Or write very efficient algorithms or do detailed security audits. But App development is just moderately complex, you can get away with basic math… So I’d say it’s doable. Still needs quite some time and effort though. At least several weeks to months. And the Kotlin book I have has like 800 pages filled with information, and that just takes some time to work through. None of it is magic, though. You do one chapter at a time.
Vibe coding is overrated IMO. There are applications and clients out there for whom it’s fine if you just do a piss-poor job and throw something together, and it somehow works enough. For a lot of things it’s not advanced enough, yet.
You don’t have to be a genius to learn programming, but it actually isn’t for everyone. Some people will never “get it” in any reasonable amount of time studying. Don’t ask me how I know!
The exact definition of vibe coding varies with who you talk to. A software dev friend of mine uses ChatGPt every day in his work and claims it saves him a ton of time. He mostly does db work and node apps right now, and I’m pretty sure the way he uses ChatGPT falls under the heading of vibe coding - using AI to generate code and then going through the code and tweaking it, saving the developer a lot of typing and grunt work.
I prefer to think of vibe coding like the relationship some famous artists had with apprentices and assistants. The master artist tells the apprentice to take care of the simple and boring stuff, like background and less significant figures. Meanwhile the master artist would paint of all the parts that require actual skill and talent. Raphael and Rembrandt would be a good examples of that sort of workflow.
My bro, your TA wasn’t better at coding because “higher IQ”. They were better because they put in the hours to build the instincts and techniques that characterize an experienced developer. As for LLM usage, my advice is to be aware of what they are and what the aren’t. They are a randomized word prediction engine trained on— among other things— all the publicly available code on the internet. This means they’ll be pretty good at solving problems that it has seen in its training set. You could use it to get things set up and maybe get something partway done, depending on how novel your idea is. An LLM cannot think or solve novel problems, and they also generally will confidently fake an answer rather than say they don’t know something, because truly, they don’t know anything. To actually make it to the finish line, you’ll almost certainly need to know how to finish it yourself, or learn how to as you go.
I vibe coded an AutoHotKey script to automate part of my job. It works.
